The New Zealand endemic Maui's dolphin (Cephalorhynchus hectori maui) is considered ‘critically endangered’ by the International Union for the Conservation of Nature as a result of decline due, in part, to fisheries-related mortalities. To estimate the abundance and trends of this subspecies, we used open-population capture–recapture models based on microsatellite genotyping of living and beachcast (dead) dolphins sampled between January 2001 and November 2007. A total of 82 genetic samples were available: 70 biopsy samples collected from living Maui's dolphins and 12 necropsy samples collected from beachcast or floating carcasses, of which five showed evidence of fisheries entanglement. Microsatellite genotyping of up to 14 loci identified 54 individuals; 42 sampled alive on one or more occasions, one sampled alive, then found beachcast 2 years later, and 11 sampled only as carcasses, including two neonates. The sex ratio of the sample did not differ significantly from unity (25 males: 29 females). Using a POPAN model for live capture records, the abundance of the super population available during the multiyear study was estimated to be N = 87 [95% (confidence limits) CL, 59–158]. Using a Pradel-like model modified to include both live capture and beachcast records, the abundance of the population was estimated to be N = 69 (95% CL, 38–125) for the midpoint of the study in 2003. The results of both models suggested that the population was likely to be declining across the study period, although this trend could not be confirmed with 95% confidence. As the genotypes provide permanent marks of individual identity, continued genetic monitoring could provide improved confidence in the abundance and trends of this subspecies.